Xiahui Li

Xiahui Li

Ph.D. Candidate in Statistics | University of St Andrews

I am a PhD candidate in the School of Mathematics and Statistics at the University of St Andrews. I hold a Master's degree in Statistics from the University of Nottingham and a Bachelor's degree in Statistics from Queen's University, Canada. My research focuses on advancing approximate Bayesian inference methods and their applications in epidemiology and ecology.

My work addresses the computational challenges inherent in likelihood-free inference for complex dynamic systems. My research interests include computational statistics, Bayesian inference, epidemiological modeling, and statistical methodology for complex stochastic systems.

Publications

Review paper figure

Advances in Approximate Bayesian Inference for Models in Epidemiology

Xiahui Li, Fergus Chadwick, Ben Swallow

Epidemics, 2025

This review synthesizes recent advances in approximate Bayesian methods that prioritize both computational efficiency and inferential accuracy. We evaluate four key families—Approximate Bayesian Computation, Bayesian Synthetic Likelihood, Integrated Nested Laplace Approximation, and Variational Inference—and provide practical guidance for their application in epidemiology.

Method paper figure

A Novel Approximate Bayesian Inference Method for Compartmental Models in Epidemiology using Stan

Xiahui Li, Fergus Chadwick, Ben Swallow

arXiv preprint, 2024

We develop a method that employs entropy criterion to select the most informative subset of summary statistics, which are then used to construct a synthetic likelihood for posterior sampling. Posterior sampling is performed using Hamiltonian Monte Carlo as implemented in the Stan software.

EDI paper figure

Actionable Guidelines for Promoting Equity, Diversity and Inclusion in Infectious Disease Modelling

Xiahui Li, Dongni Zhang, Hannah Clapham, Louise Dyson, T. Deirdre Hollingsworth, Kathy Leung, Denis Mollison, Wirichada Panngum, Ben Swallow, Cecile Tran-Kim, Jack Woodruff, Robin N. Thompson, Shweta Bansal

Epidemics, 2025 (Under Review)

This paper identifies key challenges of promoting EDI in IDD community and proposes actionable guidelines to foster inclusivity, collaboration and innovation in the field. Unlike generic EDI guidelines, we provide context-specific recommendations for three key settings in infectious disease research: team dynamics, conferences and virtual collaborations.

Controllability paper figure

Infectious Disease Outbreak Controllability: Biological, Social and Public Health Factors

Robin N. Thompson, Shweta Bansal, Hannah Clapham, Louise Dyson, Maria Gutierrez, Hans Heesterbeek, T. Deirdre Hollingsworth, Thomas House, Emily Howerton, Valerie Isham, Justin Lessler, Kathy Leung, Xiahui Li, Emma McBryde, James McCaw, Denis Mollison, Wirichada Panngum, Kris Parag, Lorenzo Pellis, Francesca Scarabel, Ben Swallow, SM Thumbi, Cecile Tran-Kiem, Cecile Viboud, Michael Plank

Proceedings of The Royal Society B, 2025 (Under Review)

This work identifies and categorizes the critical factors affecting outbreak control. We provide a structured framework, analyzing elements related to the pathogen, host population, and available interventions to guide policy assessment.

Workflow paper figure

A Workflow for Infectious Disease Modelling

Sam Abbott, Xiahui Li, Punya Alahakoon, Dhorasso Temfack, Sabine Van Elsland, Johannes Bracher, Felix Gunther, Adrian Lison, James Hay, Oliver Eales, Eben Kenah, James McCaw, Mircea T. Sofonea, Pierre Nouvellet, Freya Shearer, Sebastian Funk, Daniela De Angelis, Michael J. Plank, Anne Cori, Anne M. Presanis

arXiv preprint, 2025 (Under Submission)

We suggest a workflow for developing and evaluating infectious disease models, building on general Bayesian workflow advice and focusing on domain-specific challenges. This workflow is designed for anyone developing an infectious disease model, and for users of model outputs who need to be able to evaluate modelling studies.

Teaching

MT 1001: Introductory Mathematics

Fall 2023

This module provides a rigorous foundation in single-variable calculus, including limits, derivatives, and integrals. The focus is on building the mathematical skills necessary to model and solve problems in physics, engineering, and other scientific disciplines.

MT 2508: Statistical Inference

Spring 2024 & 2025

This module explores how we use probability models to quantify uncertainty and make inferences from data. Students learn key statistical methods, including maximum likelihood estimation, confidence intervals, hypothesis testing, and linear regression, forming the essential foundation for advanced statistical study.

MT 2501: Linear Mathematics

Fall 2024

This module builds on matrices and linear systems to explore the fundamental structures of linear algebra: vector spaces, linear independence, transformations, and diagonalization, with applications across the mathematical sciences.

MT 4113: Computing in Statistics

Fall 2023, 2024 & 2025

This module builds practical programming skills for statistical analysis. Students learn to write efficient, modular R code for data manipulation, simulation, and investigating statistical procedures.

Summer Team Enterprise Program Coach

Summer 2024

Coached a team of undergraduates through exploratory data analysis of oceanographic datasets, from research question formulation to final reporting. Their insights directly contributed to the development of the new module, "The History and Future of Data."

Presentations

Conference & Workshop Talks

Approximate Bayesian Inference For Models in Epidemiology
International Statistical Institute World Statistics Congress, The Hague, Netherlands
October 2025 • Slides
Approximate Bayesian Inference Using Policy-informed Summary Statistics
International Center for Mathematical Sciences, Edinburgh, UK
October 2024
A Novel Approximate Bayesian Inference Method for Compartmental Models in Epidemiology Using Stan
Isaac Newton Institute for Mathematical Sciences, Cambridge, UK
August 2024 • Video

Invited Seminars

Actionable Guidelines for Promoting Equity, Diversity and Inclusion in Infectious Disease Modelling
JUNIPER Virtual Seminar Series
November 2025 • Slides
A Novel Approximate Bayesian Inference Method for Compartmental Models in Epidemiology Using Stan
JUNIPER Virtual Seminar Series
January 2025 • Slides

Curriculum Vitae

Education

Research Activities

World Statistics Congress 2025
World Statistics Congress October 2025
International Statistical Institute, The Hague, Netherlands
CIRM Workshop 2025
International Workshop on Analysis and Modelling for the Design of Future Epidemic Surveillance Systems April 2025
Centre International De Rencontres Mathématiques, Marseille, France
ICMS Workshop 2024
Workshop on Mathematical Models for Twenty-First Century Decisions October 2024
International Center for Mathematical Sciences, Edinburgh, UK
Isaac Newton Institute 2024
International Programme on Modeling and Inference for Pandemic Preparedness August 2024
Isaac Newton Institute, Cambridge, UK

Professional Affiliations

PhD Student Researcher 2023 – Present
Machine Learning and Statistics Methods Research Group, University of St Andrews
Researcher 2024 – Present
Joint Universities Pandemic and Epidemiological Research (JUNIPER) Consortium
Associate Fellow 2024 – Present
Graduate School for Interdisciplinary Studies, University of St Andrews

Honors & Awards

Presentation Award 2025
International Statistical Institute World Statistics Congress
St Leonard's College Postgraduate Travel Award 2025
University of St Andrews
Durham Summer Grant 2024
International Workshop on Statistical Modelling
PhD Scholarship 2023 – 2027
China Scholarship Council and University of St Andrews
Nottingham Advantage Award 2021 – 2022
University of Nottingham
Download Full CV (PDF)